08 January 2013

Game Theories: Epic Fails

I was having a late-night discussion with some designers at work about the nature of Succeeding and Failing in games. The question of the hour was this:

"In a narrative-driven or narrative-inflected game, what is a player's threshold for tolerating narrative failures with respect to gameplay."

To illustrate this, consider an exemplary assassination set-up in a typical Assassin's Creed game: You are given a target, and discover multiple approaches to reaching him. After choosing your path, you sneak up on the target and press the final "Attack" button. The assassination animation kicks in ... then cuts to one of the following cinematics:

1. You plunge your blade into the target's chest, killing him. [Success!]

2. You plunge your blade into the target's chest, but the target is wearing an iron vest a la the Man With No Name in "A Fistful of Dollars." Surprise! Your blade breaks. He has tricked you. After you wrestle, your target escapes. [Failure!]

3. You slam your fist down, but your target slips to one side at the last moment. The blade misses completely and slams into the ground. Your target rolls away and escapes. [Failure!]

Two of these scenarios end in narrative failure, one in success, and yet, in all three the gameplay loop is exactly the same -- the planning, approach, the execution of the attack. In all three cases, the player can be said to have "played the same game."

 Now, my colleagues felt that -- of all options -- two is slightly frustrating, but dramatically acceptable, while three is wholly unacceptable. You cannot "force fail" the player without there being some outwardly apparent reason. In option 3, one would assume the man avoided your attack because he is skilled and swift; but this did not satisfy my colleagues.

I maintained that this wasn't exactly a force-fail, since the player actually carried out the complete game-play loop required to "win." Yes, the narrative outcome is frustrating, but only in a story sense. And if the story's intent is to frustrate the player (after all, art can and should be provocative sometimes) then this might be a fine way to accomplish it. So I put their aversion to this sort of failure down to a question of taste and tolerance. I could be terribly wrong.

Admittedly, option 3 is a bummer. But is it inherently worse than option 1, even if it clearly serves the story in some way? Is option 3 not simply an experimental form of narrative -- a deliberately provocative or subversive narrative attached to proven gameplay loops? The question remains open.

I suspect a large number of people would still say that option 3 is terrible, even if, in all three cases, there is not a jot of difference in the gameplay. This is the puzzling thing about narrative in games. The context it provides to what you are actually DOING -- killing people rather than pressing X buttons to interact with rectangles -- is very alluring. So despite the player in our example executing the same task in all three cases, he will feel quite differently about them depending on that final cutscene.

One extreme variation we agreed was forbidden was the "forced gameplay fail." This could come about by forcing the player into a position where they were made to execute exactly the opposite action they intended at the start of the loop. In such cases, the game IS cheating you by forcing actions you have been deliberately avoiding.

So let's say you reach your target, and are unable to actually "lock on" to him. Instead you are forced to get so close that a cutscene automatically triggers. Whether or not you kill the target in the cutscene, it doesn't matter. The gameplay has robbed you of the final decision, the final strike. This is a forced gameplay-fail and such malarky should be avoided at all cost.


  1. Part of the issue at hand is that we're not talking about a strict simulation, as in, say, Gran Turismo. If I hit a curve at 85 MPH and brake for two seconds, I should get the same result as the last time I did the same thing.

    Pressing the Attack button is the narrative equivalent of the "If you want to try to attack that guy, go to page 37." Emphasis on the "try" in this case. I can see how if this is the first time pressing that button results in a miss it could come as a shock, but moving the story forward should win out.

  2. A good point. However, I suppose my colleagues would argue that the attack button SHOULD function like the brakes in GT, so long as all the external conditions have been met.

    So IF I am incognito, and IF I am standing above my target, and IF I hit the attack button, then the output should be knowable. That's because the button, in this case, means "Air Assassinate". It does 99 percent of the time, so why not in this one case? I'm not sure I agree, but I suspect that's what they'd say.

    In the context of open-conflict combat of course, NOW the button means "Swing your Weapon" ... and if the conditions are right, you will land a blow. But we are aware that our enemy may block the blow, because the awareness of the enemy offers a host of new external conditions to keep track of.

  3. I think that all of these options technically would be acceptable. If anything, option three is a dulled down version of option two. If the attack fails then why not honour the title of the entry and make it an epic fail? Right now, if option three is chosen then the character misses their target but is otherwise in the same place that they were beforehand. Why not get him into trouble?? This is why I really like option two. A character has made a bold choice to assassinate their target at this specific moment. There should be a risk that goes with that choice. After the target escapes in this option, the character has failed and he has lost the use of his primary weapon (a weapon that has probably been heavily relied on up to this point). This not only pushes the player onward but also pushes them out of their comfort zone. This will make the eventual success far more rewarding for the player, especially since they will be forced to exploit more of the game’s creativity in their second approach in order to succeed.

    But aside from the options themselves, I think that the context of this event is what will really make an action acceptable from the perspective of the player. From a writing perspective it comes down to setups and payoffs. If the player is given information or backstory on the target that makes one of the three options easier to buy into, then they’ll accept it, but if the outcome comes completely out of the blue, then they’ll have a problem no matter how legitimate the action is from a technical standpoint.